Hosting annual international sporting events and tourism: Formula 1,golf or tennis?

Abstract

Hosting sports events to attract international tourists is a common policy practised by many host governments. Hosting mega-sports events like the Olympics is said to leave a legacy that could impact the attractiveness of a country/city in the long term. However, the opportunity to host these mega-events is limited and expensive. This study considers the economic impact of hosting annual international sporting events, specifically the extent to which Formula 1, ATP Tennis and PGA Golf can attract international tourists. Using monthly data from 1998 to 2018, we show that the effect differs from one sport to another within a country and the same sport across countries. Hosting the Formula 1 is most effective for Canada but has no significant impact in Australia and the United Kingdom. ATP Tennis and PGA Golf have a significant impact on at least two countries. Policy-makers must consider carefully the sport that gives the best bang-for-the-buck.

Keywords

sports tourism Formula 1 tennis golf international tourists

Introduction

Hosting sporting events to attract more international tourists has become a popular strategy among many local and national governments. The Brazilian Ministry of Sports estimated that the hosting of the 2014 FIFA World Cup would attract 600,000 international tourists while the previous tournament in South Africa attracted more than 300,000 foreign visitors (Baumann and Matheson, 2018). Hosting large-scale, high-profile sports events has been on the increase in recent years as there is a strong belief that there is a net positive economic impact from hosting such events (Huang et al., 2014), particularly when the expected CAGR of international sports tourism over the next 7 years is 32.3% (ReportLinker.com). The city of Shanghai, for example, is reported to host about 160 sports events a year including the Formula 1, men’s and women’s golf and the Shanghai Masters Tennis in its effort to become a sports metropolis by 2025, emulating other well-known sports destinations like London and Paris (South China Morning Post, 29 August 2018).

The rationale for hosting large-scale sporting events is that the event will leave a legacy such that the impact of the event will ‘remain longer than the event itself’ (Preuss, 2007: 211). However, Thomson et al. (2019) state that a legacy is created only if the scale of the event involves ‘significant investment in infrastructure and urban development, have international media exposure and attract large number of tourists’ (p. 295). Clearly, these conditions refer to mega-events like the Olympic Games and the FIFA World Cup. However, opportunities to host these mega-events are limited to a handful of cities/countries within a decade that are capable of meeting a long list of criteria. In recent years, the net benefits of hosting such events have been questioned (Zimbalist, 2016) as to whether it is a fool’s gold (Baade and Matheson, 2002) or a lottery jackpot (Preuss, 2006: 183).

In this article, we turn away from mega sporting events which dominate the current literature and focus our attention on popular annual international sporting events. We focus on Formula 1 Grand Prix, ATP Tennis tournaments and PGA Golf played in Australia, Canada and the United Kingdom. These sports are among the top 10 biggest global sports¹ based on coverage of major online sports websites and that follow an annual international schedule/circuit. They are hosted by the same city/country annually in a particular month and so allows the city/country to make the event a permanent feature in its calendar and over the long term, link the image of the sport with the destination (for instance, cycling and France or Wimbledon and tennis). We assume the popularity of these global sports often translate into tourist attractions and evaluate the potential economic impact of hosting these annual events, that is, the increase in international tourist inflow. We compare the three sports within the country as well as each sport across the three countries. We propose an inclusive framework in this article where other sports can be added and evaluated.

The effect of the COVID-19 pandemic on the tourism industry is unprecedented. Behsudi (2020) reported that worldwide tourism receipts are not expected to recover to pre-pandemic levels until 2023. Since the hospitality and travel industry make up on average 10% of employment and 9.5% of GDP among the G20 countries, the revival of this industry will be critical for the general recovery of the world economy. Although in the short term, the focus of governments has been on providing grants and debt reliefs to enterprises and corporations in the industry, in the longer term, efforts have to be made to jump start the tourism industry. The findings of this study could assist policy-makers in determining those annual sporting events which could attract the international tourist.

Literature review

Gunn (1972) affirmed that without attractions, there will be no tourists. Hu and Wall (2005) further stated that tourist attractions are essential ingredients to develop a successful tourist destination. They further define an attraction to be a permanent feature that is either natural or man-made, as long as the main purpose of its development is to attract tourists. Hall (1992) identified three domains in sports tourism – hallmark events, health and fitness and outdoor recreation. Hallmark events can be seen to be the most direct and substantial link between sports and tourism (Hinch and Higham, 2001). Given the growth in sports tourism in recent years due to the increasing popularity of sports in general, as well as active lifestyles in particular (Green and Chalip, 1998; Roche et al., 2013), hosting popular sporting events that attract enthusiasts, both domestic and foreign, is an obvious strategy among many city governments.

Previous literature that evaluates the economic impact of hosting hallmark sporting events have focused predominantly on mega sporting events (see for example, Baumann and Matheson, 2018; Mitchell and Stewart, 2015; Rojas-Méndez et al., 2019, among others). Ritchie (1984) revealed various ways in which these events can make an impact, potentially resulting in the so-called hallmark events. Using a gravity model involving 200 countries over the period 1995 to 2006, Fourie and Santana-Gallego (2011) showed that hosting these mega-events can increase tourist arrivals by about 8% a year. However, Mitchell and Stewart (2015) found that the Beijing Olympics did not have any significant effect on Chinese tourism, perhaps due to the crowding out effect raised by Baumann and Matheson (2018), where the negative externality due to the congestion of a mega-event is likely to dissuade tourists to travel to the host region. More recently, cities like Boston, Budapest, Rome and Hamburg had to withdraw their bidding for the 2024 Olympic Games due to a backlash from their residents (Thomson et al., 2019). In general, the results are mixed depending on whether the studies are prospective or retrospective (Humphreys, 2006). These mega-events are however a once-in-a-life-time event, and the pre-requisites and competition to host them are challenging. FIFA, for example, requires host countries to have between 8 to 12 large stadiums, while the Olympic International Committee require facilities that can house 15,000 athletes and sufficient hotel rooms for spectators (Baade and Matheson, 2002).

Several studies have considered the impact of smaller, local but regular sporting events in the United States (Daniels and Norman, 2003; Kaplanidou et al., 2012) and Japan (Nogawa et al., 1996), while several others have evaluated the economic impact of a single sport (e.g. Kim et al., 2017 and Ramasamy and Yeung, 2020 for Formula 1; Papanikos, 2015 for the Athens Marathon). A majority of the literature examines the perception of the destination image (Hallmann et al., 2015; Kaplanidou et al., 2012; Kaplanidou and Gibson, 2010), visit motivation (Yusof et al., 2009) and focus on tourist profiling (Yusof et al., 2009; Ziakas and Boukas, 2016). With an abundance of research on mega-events and to a certain degree, small-scale or individual events, we find limited research that takes a closer look at the impact of hosting annual international sporting events, let alone comparing which annual international sporting event might be more beneficial for the hosting city/country. Huang et al. (2014) is a rare paper that compares three major sports events in China to estimate the likely economic impact. However, they use survey data and include the spending of locals, which muddles the results. In this article, our focus is on popular international sporting events that occur on an annual basis, namely, the Formula 1 Grand Prix, PGA Golf and ATP Tennis. Our article bridges this gap in the literature. We hope to contribute to the policy conversation and widen the possible events for countries that are less likely to host mega-events yet still wish to attract international tourists who are sports enthusiasts. In addition, unlike most papers in the literature that use case studies and questionnaires, which leads to relatively limited data for more general implications, our article instead focuses on one aspect that would contribute to the potential economic benefits of hosting different annual international events, namely, the number of tourist arrivals.

There are two important reasons why hosting annual international sporting events may be preferable to a single mega-event. First, the financial burden of hosting the Olympics or the FIFA World Cup may not be affordable by many countries/cities given the requirement mentioned earlier (Gibson et al., 2012; Könecke et al., 2017; Solberg and Preuss, 2007). The Australian government is reported to have committed US$46 million just to bid for the 2022 World Cup and was prepared to spend a further US$3 billion to build infrastructure and stadiums if successful (Mitchel and Stewart, 2015). Gruben et al. (2012) reported that London spent US$25.5 million just to win the bid to host the 2012 Olympic Games and a further US$4 billion to host the Games. In contrast, Sylt (2017) estimated the average cost of hosting the Formula 1 races to be under US$60 million annually (in addition to a one-time cost of US$270 million for building the race-track). Wan and Song (2019) find that developed and developing countries consider these expenditures differently. While the 2012 London Olympics is said to boost the British economy in general, Brazil’s reason was to promote tourism but at the cost of local development. Schulz (2010) pointed out that mega-events are usually limited to a 3-5 year build-up phase, a two- to six-week event and then follow-up attempts to leverage legacies. Recurring sports events on the other hand are firstly repetitive and often are based on existing local infrastructure and have more potential to evolve with local opportunities and needs. Giampiccoli et al. (2015) assess the difference between the 2010 World Cup and recurring sporting events and suggest that a focus on the latter will yield more sustainable and predictable returns that benefits host communities.

Second, mega-events happen once in several decades. For example, the host of the 32nd Summer Olympics 2020 in 2021, Tokyo, also hosted the 18th Olympiad in 1964 – a gap of more than 50 years. Such long-term economic impact of hosting these mega-events is doubtful. Since these events are flashpoints in history, the tendency to forget previous hosts is high (Solberg and Preuss, 2007). One year after the hosting of the European football tournament (EURO 2000), 55% of survey respondents forgot who the hosts were (Oldenbloom, 2006). Annual sporting events, on the other hand, have the capability of converging the event and sports image to the destination image and create a sustainable pool of repeat visitors over a longer term (Kaplanidou et al., 2012; Taks et al., 2015). Daniels and Norman (2003) show that regular sporting events can provide significant economic potentials for the host, especially when a combination of several events over the year is carried out.

Governments and/or national sports associations have a choice as to which major sporting event that they can bid for. These sports differ in terms of their size and significance; thus, the ability of the sport to attract tourists and media coverage would vary accordingly (Getz, 1997). It is important that stakeholders understand the economic impact of hosting various sporting events and choose those events that offer the highest returns (Huang et al., 2014). More specifically, in this article, we compare the economic impact of hosting the selected sports, in particular the ability of these sports to attract international tourists. Expenditure by tourists makes up a significant portion of the new money that flows into a city/country which leads to the creation of new jobs and income (Crompton et al., 2001; Mitchell and Stewart, 2015). In the case of China, for example, Huang et al. (2014) found that only 12% of the new money from hosting the Formula 1 races came from local attendees. Thus, the ability of a particular sport to attract international tourists can represent a significant aspect of the potential economic impact of playing host for the event. Realizing this, China is planning the nation’s first national sports tourism pilot zone in Hainan Island, aiming to host international sports events more frequently (China Daily Global Edition, 9 April 2020).

Five countries host the three sports selected for this study at a high level – Australia, Canada, the United Kingdom, France and Spain. Our choice of the former three is to ensure a geographical spread across various continents. All three selected host countries are, however, part of the British Commonwealth but more importantly have a long history of hosting these sports. Some brief details of these sports tournaments at various locations are described in Table 1.

Table 1.

Tennis, Formula 1 and golf in Australia, Canada and the United Kingdom: Some basic facts.

The Association of Tennis Professionals (ATP) is the governing body for tennis, and the ATP Tour is the main tennis tournament played across several cities around the world. The ATP also has a system of ranking players based on the performance of the player at 19 specific tournaments. Among the popular tournaments include the Grand Slams and the ATP Tour Masters 1000. The tournaments considered in this study are the Australian Open, Wimbledon (both are part of the Grand Slam) and the Canadian Open (one of eight mandatory tournaments that accrue points for the world rankings). The Australian Open, which is normally played in January in Melbourne, was first held in 1905. The Wimbledon was inaugurated in 1877 and is normally played in June and July. The Canadian Open (also known as the Rogers Cup) goes as far back as 1881. The tournament is usually played in August and is alternated between Montreal and Toronto. In 2019, Novak Djokovic won the Australian Open and Wimbledon while Rafael Nadal won the Canadian Open. In the women’s tournament, Naomi Osaka, Simona Halep and Bianca Andreescu won the Australian, Wimbledon and Canadian Open, respectively. Tennis ranked as the third biggest global sport and is a major sport in all three countries in our sample.

Formula 1 racing started in 1950 and has become the foremost motor racing competition, taking place in 22 countries, either on dedicated circuits or on public roads. The races are sanctioned by the Fédération Internationale de l'Automobile (FIA) and are owned by the Formula One Group. The F1 calendar starts in March and ends in December. The Australian GP in Melbourne usually starts the year, with the Canadian GP taking place in June and the British GP in July. All three races are held in dedicated circuits. All three Grand Prix have a long history in their respective countries. The British, Canadian and Australian GPs started in 1950, 1967 and 1985, respectively. In 2019, the winner of the British and Canadian GPs was Lewis Hamilton (Mercedes) while Valtteri Bottas (Mercedes) won the Australian GP. F1 ranked as the fifth biggest global sport and is a minor sport in Australia and Canada but considered major in the United Kingdom.

The Australian Open is the most prestigious golf tournament in Australia. It is part of the PGA Tour of Australasia and has been played since 1904. Although several cities have hosted this tournament, Sydney and Melbourne have been the most popular hosts. Founded in 1860, the British Open is the oldest golf tournament in the world. Played usually in July, the tournament is rotated around several golf courses in the United Kingdom. The Canadian Open is co-organized by Golf Canada and the PGA Tour and has a history that begins in 1904. Although played at various courses in Canada, Ontario has hosted most of this tournament in recent years. In 2019, Matt Jones, Shane Cowry and Rory McIlroy won the Australian, British and Canadian Open, respectively. Golf ranked as the tenth biggest global sport and is a major sport in United Kingdom, while medium in the other two countries.

Sources: www.atptour.com; www.formula1.com; www.pgatour.com; www.biggestglobalsports.com.

Data and methodology

Previous studies have used two popular approaches to model tourism demand at the country level. The first approach, the Box-Jenkins univariate model, features the time-series of tourism demand itself as an exogenous variable. The rationale here is that without any other exogenous variables, the model is able to detect a great deal of underlying behaviour within the tourism demand time-series (e.g. the time trend, moving average structures and autoregressive parameters). To determine if an event had a significant impact on tourism, dummy variables representing the timing of the event are appended to the time-series model (i.e. ARIMA). Mitchell and Stewart (2015), for example, fitted an autoregressive moving average model with a linear deterministic trend, monthly dummy variables to account for the time trend and other dummy variables to account for changes in tourism demand slumps due to catastrophes like SARS. A dummy 0 was included for the period before the event and a 1 after, to account for the change in the slope of the trend. The model is a pure univariate model without including any other exogenous variable.

The second approach, which has been used in this study, models the variations in tourism demand by regressing it on a few selected exogenous variables. There have been many attempts to explain the variations in inbound arrivals of a particular country using a range of regressors. For instance, Witt and Witt (1995) suggested using the lagged dependent variable, which has been widely followed in the literature. Song et al. (2010: 73) stated that the ‘lagged dependent variable describes tourists’ expectations, habit persistence, the “word-of-mouth” effect and supply constraints’. Time trend variables are included to represent tourists’ changing tastes and capture other time-dependent effects (Witt and Witt, 1995). Political stability has been shown to influence attendance at the Olympic Games (Gruben et al., 2012). Seetaram (2010) showed that the population of overseas Australians living in the source country was strongly related to inbound tourism to Australia. Kusni et al. (2013) highlighted the importance of relative price and substitute price (between the host and competing locations) as important determinants of tourist arrivals. These studies show that the selection of regressors and the model specifications tend to be country-specific; hence, there has been little consensus among researchers as to what constitutes a standard reference model. In fact, more than 95% of studies used models with dyadic data which addressed issues for a single country (Song and Li, 2008). A few isolated efforts have been made to guide the identification of essential elements for modelling the variations in tourism demand. Pham et al. (2017) reviewed tourism demand studies over the past five decades and suggest four essential features for modelling the variations in tourism demand: (1) a variable that denotes the level of stability of the destination; (2) a variable representing the attractive feature of the destination; (3) a variable representing the travelling cost to the destination and (4) lagged variables representing expectations, habit persistence, the ‘word-of-mouth’ effect and supply constraints. Ramasamy and Yeung (2020) applied the same framework and found the effects of hosting the Formula 1 Grand Prix on tourism demand to be significant but stressed that there is neither an impeccably well-specified model nor a fixed selection of standard regressors for modelling tourism demand.

Nielsen Sports (2018) surveyed 18 countries across North and South America, Europe, the Middle East and Asia and concluded that the 10 most popular sports based on interest were football (soccer), basketball, athletics, tennis, motorsports, cycling, cricket, mixed martial arts (MMA), baseball and golf. Although all these sports have some form of competitive events at a global level, we identified tennis, motorsports and golf as globally popular sports that have an annual calendar with competitions taking place in permanent locations (cities). More specifically, in this article, we compare the effect of hosting the Formula 1 races, PGA Golf and ATP Tennis on tourist arrivals in Australia, Canada and the United Kingdom. Since these events take place during a particular month during the year, we used monthly data to study the variations in arrivals. Moreover, since most of these sporting events have lead-up days or practice days, monthly data would be able to capture visitors and participants arriving before as well as those who may linger after such events. Monthly data for the common variables used in previous studies explained above are either non-existent or difficult to access. Nevertheless, we have attempted to include proxy variables that would best meet the requirements as stated in Pham et al. (2017). For the current study, we include the following:

The dependent variable, TOURISTS_t, is the number of inbound international tourist to country i (Australia, Canada and the United Kingdom) in month.

The word-of-mouth effect is measured by the number of tourists received in previous periods (Chaisumpunsakul and Pholphirul, 2018; Fu et al., 2020; Kusni et al., 2013; Pham et al., 2017). LAGN_(t−1) and LAGS_(t−11) measure non-seasonal and periodic lags to proxy habit persistence.²

To account for the cost of living in the destination country, we use the real effective exchange rate of destination countries (EFFEX_t) (Görmüş and Göçer, 2010). This measure consolidates the relative price difference between the destination country and the home country used by some studies (Allen and Yap, 2009; Chaisumpunsakul and Pholphirul, 2018; Kosnan and Ismail, 2012; Kusni et al., 2013; Naudé and Saayman, 2005; Witt and Witt, 1995) and the nominal exchange rate by others (Kosnan and Ismail, 2012).

The MSCI World Index (MSCI_t) is a weighted stock market index that includes significant companies worldwide. Most papers in the literature (Martins et al., 2017; Pham et al., 2017; Seetaram, 2010) used GDP per capita to measure the global economy. However, GDP data are only available quarterly, whereas the MSCI World Index offers data at a higher frequency. We use the MSCI index reported on the last trading day of the month to capture the global economic condition at a monthly frequency in our model.

The number of reported terrorist attacks (both actual and potential) in the respective country in month t reported by popular media (TERRORISM_t) is used to proxy the stability of the destination. We used University of Maryland’s Global Terrorism Database to measure this control variable.

We also include temperature in the destination country in month t (TEMP_t and TEMP_t²) as a continuous variable.³ Previous studies have used a dummy variable to control for seasons (Allen and Yap, 2009; Fourie and Santana-Gallego, 2011). Since the three sporting events may occur in different months/seasons, using the average monthly temperature can offer more insights without assuming any particular seasonal peak for tourists’ arrival.

It is likely that a country may also host events, other than the three sports, during a particular month. We included EVENTS_t to take into account various international events that the country plays host in month t. EVENTS_t is a count of the number of news articles from a popular daily in country i in month t that carried the terms ‘international’, ‘sports’, ‘exhibition’ and ‘festival.’⁴ The dailies chosen, based on their popularity, were the Herald Sun, Toronto Stars and the Daily Mail for Australia, Canada and the United Kingdom, respectively.⁵

Dummies for the years 2003 and 2008 are included to account for two major events that affected tourism worldwide, namely, the SARS epidemic and the Global Financial Crisis (FIN), respectively. Following other tourism demand studies (Witt and Witt, 1995), a time trend variable (TREND) is included to represent tourists’ changing tastes and to capture other time-dependent effects, including changes in the ease and cost of travel over time.

Finally, the variables of interest, namely, F1, TENNIS and GOLF representing the number of days that each sporting event was hosted by each country in month t are included. If the sporting event was held across 2 months (e.g. the case of the Wimbledon), the variable would have two non-zero entries. If more than one sporting event was held in the same month (e.g. in the case of the United Kingdom), the number of days for each sport in that month is included in the respective sport variable.

Our data span the period from January 1995 to December 2018. Data sources for the above variables are listed in Table 2. The model described above can be written as shown below and fitted with monthly data for the three countries of interest

\begin{array}{l} {TOURISTS}_{(t)} = β_{0} + δ_{1} {TENNIS}_{(t)} + δ_{2} {GOLF}_{(t)} + δ_{3} F 1_{(t)} \\ + β_{1} {LAGN}_{(t - 1)} + β_{2} {LAGS}_{(t - 11)} + β_{3} {EFFEX}_{(t)} + β_{4} {MSCI}_{(t)} \\ + β_{5} {TEMP}_{(t)} + β_{6} {TEMP}_{(t)}^{2} + β_{7} {TERRORISM}_{(t)} + β_{8} {EVENTS}_{(t)} \\ + β_{9} {SARS}_{2003} + β_{10} {FIN}_{2008} + β_{11} TREND + ε_{(t)} \end{array}

Table 2.

Data sources.

Variable	Canada	UK	AUS
${TOURISTS}_{(t)}$	Statistics Canada https://www150.statcan.gc.ca/n1/en/subjects/travel_and_tourism	UK Office for National Statistics https://www.ons.gov.uk/peoplepopulationandcommunity/leisureandtourism	Australia Bureau of Statistics https://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/3401.0Mar2020?OpenDocument
${EFFEX}_{(t)}$	Bank for International Settlements https://www.bis.org/statistics/eer.htm
${MSCI}_{(t)}$	Thomson Reuters’s Datastream
${TEMP}_{(t)}$	Data portal of Canadian Centre for Climate Services www.climate data.ca	Met Office The national meteorological service for the United Kingdom www.metoffice.gov.uk	Bureau of Meteorology, Australia www.bom.gov.au
${TERRORISM}_{(t)}$	Global Terrorism Database from The National Consortium for the Study of Terrorism and Responses to Terrorism, University of Maryland, https://www.start.umd.edu/data-tools/global-terrorism-database-gtd
EVENTS_(i)	The frequency counts of news articles in a major daily in country i that carried the terms ‘international’, ‘sports’, ‘exhibition’ and ‘festival’. The dailies were Herald Sun, Toronto Stars and the Daily Mail for Australia, Canada and the United Kingdom, respectively.
$F 1_{(t)}$	www.f1.com
${GOLF}_{(t)}$	PGA of Canada www.pgaofcanada.com	PGA www.pga.info	PGA of Australia www.pga.org.au
${TENNIS}_{(t)}$	Rogers Cub www.rogerscup.com	Wimbledon www.wimbledon.com	Australia Open www.ausopen.com

All continuous variables in the model are log-transformed. We examined the possibility of employing other functional forms. The commonly used log–log, log-linear and linear forms were all considered. The possible variations in the coefficient of the sports variables across functional forms were examined. We found that the significance of the coefficients estimated for the three sports events across the three models were quite consistent. However, the log–log model provided the most desirable statistical properties in terms of the lowest AIC value and an insignificant Jarque–Bera statistic. We concluded that the log–log specification is appropriate as it imbues the fitted models with a high level of interpretability and therefore is used to guide the rest of the analysis.

The model described above allows us to compare the effectiveness of a particular sporting event to attract inbound tourists to each of the three selected countries. However, we are also interested in comparing the ability of the sport to attract tourists across countries. In other words, we wish to analyse the strength of the sporting event coefficients ( $δ_{1}$ , $δ_{2}$ and $δ_{3}$ ) across equations so that comparisons can be realistically carried out, if all equations are estimated under a system of equations. Seemingly unrelated regression (SUR) is able to show the variations in the relationships across the data dimensions as well as provide a convenient vehicle for testing our hypotheses (Fiebig, 2003). The correlations among error terms of the equations represent common shocks⁶ and to take into account these features, a correct estimator must be applied to produce unbiased and consistent results. Since we are examining the tourism demand of all three countries within the same time frame, contemporaneous correlations among the disturbances must be considered, and the common approach for doing this is referred to as seemingly unrelated regression estimation (SURE), developed by Zellner (1962). The remedial treatment derived from Aitken’s generalized least square modified by Wang et al. (1980) is used to estimate our system of equations. In this approach, the inclusion of lagged dependent variables becomes possible, with an additional advantage of reduced serial correlation. When analysing tourism demand with a system of demand equations specifically for different countries, the potential correlations of the disturbances across these equations may arise because a shock affecting the demand in one location may spill over and affect the demand of other countries. To account for such cross-equation contemporaneous correlations of the disturbances, SURE will result in smaller standard errors of the estimates and hence produce more precise estimates of the coefficients.

Results

The results of our estimations using the OLS and SURE methods are reported in Tables 3 and 4, respectively. In the OLS estimation, we detected serial correlations when lagged dependent variables were included. Driscoll–Kraay standard errors that are robust to heteroskedasticity and serial correlation up to several lags were computed to address these.

Table 3.

Estimates using OLS log–log (original).

	Aus		UK		Canada
	Coeff.	p-value	Coeff.	p-value	Coeff.	p-value
Constant	5.169***	0.000	3.789***	0.000	10.960***	0.000
${TENNIS}_{(t)}$	−0.006	0.879	0.006**	0.038	0.006*	0.069
${GOLF}_{(t)}$	0.021***	0.001	0.042***	0.000	−0.003	0.625
$F 1_{(t)}$	0.003	0.797	0.035	0.434	0.129**	0.026
${LAGN}_{(t - 1)}$	0.417***	0.000	0.292***	0.000	0.372***	0.000
${LAGS}_{(t - 11)}$	0.339***	0.000	0.064	0.277	−0.219***	0.000
${EFFEX}_{(t)}$	−0.001	0.629	−0.003**	0.024	−0.006***	0.000
${MSCI}_{(t)}$	0.011	0.837	0.088	0.104	0.172***	0.000
${TEMP}_{(t)}$	−0.261***	0.000	0.010	0.285	0.028***	0.000
$TEMP 2_{(t)}$	0.008***	0.000	−0.001	0.214	0.001***	0.000
${TERRORISM}_{(t)}$	−0.001	0.988	−0.001	0.949	−0.040*	0.054
EVENTS_(t)	0.006***	0.027	0.009***	0.000	−0.003	0.101
${SARS}_{2003}$	−0.300	0.611	−0.241***	0.000	−0.206	0.118
${FIN}_{2008}$	−0.008	0.812	−0.170***	0.000	−0.024	0.557
TREND	0.001**	0.046	0.001	0.000	−0.001**	0.021
						AIC<−1
R²	0.89		0.88		0.83	JB=2.03

The stationarity of the data series was examined by applying the usual unit root tests. There was significant evidence to reject the unit root hypothesis, including the MSCI. Results are available upon request. The Newey–West method correcting for autocorrelation heterogeneity of variances in the errors were applied. Observing PP- and QQ-plots, despite a minor and trivial deviation from normality for the Canadian model, we concluded that the residuals are close to a normal distribution. Nevertheless, the Jarque–Bera statistics (2.03) on the residuals for the Canadian model was not rejected.

***, **, * indicate significance at the 1%, 5% and 10% levels, respectively.

Table 4.

Estimates using SURE.

	Aus		UK		Canada
	Coeff.	p-value	Coeff.	p-value	Coeff.	p-value
Constant	5.117***	0.000	3.464***	0.000	10.664***	0.000
${TENNIS}_{(t)}$	−0.003	0.891	0.004**	0.043	0.007***	0.005
${GOLF}_{(t)}$	0.020***	0.000	0.042***	0.000	−0.009	0.197
$F 1_{(t)}$	0.002	0.794	0.026	0.350	0.143***	0.000
${LAGN}_{(t - 1)}$	0.417***	0.000	0.312***	0.000	0.381***	0.000
${LAGS}_{(t - 11)}$	0.332***	0.000	0.094	0.114	0.203***	0.000
${EFFEX}_{(t)}$	−0.001	0.585	−0.003**	0.020	−0.006***	0.000
${MSCI}_{(t)}$	0.023	0.636	0.085	0.108	0.165***	0.001
${TEMP}_{(t)}$	−0.257***	0.000	0.016*	0.062	0.018***	0.000
${TEMP 2}_{(t)}$	0.008***	0.000	−0.001**	0.043	0.001***	0.000
${TERRORISM}_{(t)}$	0.004	0.812	−0.001	0.982	−0.068***	0.001
EVENTS_(t)	0.007***	0.006	0.010***	0.000	0.003**	0.038
${SARS}_{2003}$	−0.340	0.305	−0.154***	0.000	−0.250***	0.000
${FIN}_{2008}$	−0.008	0.761	−0.254***	0.000	−0.025	0.564
TREND	0.001**	0.012	0.001***	0.001	−0.001	0.309

An R² for Aitken’s generalized least square model is not computed here since the replacement pseudo R² is also not well defined.

***, **, * indicate significance at the 1%, 5% and 10% levels, respectively.

As shown in Table 3, our results are markedly different for each of the country considered in this analysis. In the case of Australia, the lagged variables are strongly significant indicating the influence of word-of-mouth effects on tourist inflow into the country. A warmer climate is generally preferred among tourists. International events in general help attract international tourists to Australia. The level of wealth, cost of living and world crises like SARS and the 2008 Financial Crisis are not significant in our model. As for sports, only PGA Golf has a significant positive impact on international tourist inflow. In contrast to Australia, our UK results point to the importance of cost-of-living differences and world crises that may have a negative effect on international tourists. The lagged dependent variable and international events are significant and positive. Two sports event – ATP Tennis and PGA Golf – have a significant effect on tourist inflow in the United Kingdom.

More control variables are significant in the Canadian model. In fact, the variables that are not significant are the two world crises and the international events proxy. As for the sports events, in Canada, the F1 and ATP Tennis are strong attractors of international tourists.

Comparing Tables 3 and 4, OLS and SURE estimates are quite consistent for most of our control variables. In addition, for the sports variables, both models also produce similar results for all three countries, but with a greater degree of significance. The OLS estimates are obtained while ignoring any correlation between the error terms across equations. However, if the error terms are contemporaneously correlated, as is most likely in the case of tourism demand studies, the estimation procedure should take this into account. In this case, the SURE estimator leads to more efficient parameter estimates.

Several control variables deserve special mention. First, the temperature variable (TEMP) is significant for Australia and Canada. In both cases, computing the total effect of the two temperature variables generally results in a positive relationship, that is, warmer climates are preferred by tourists in both countries. It is also possible that the warmer temperature coincides better with the school calendar for travelling. Second, the TERRORISM variable is significantly negative only in Canada. Third, the EVENTS variable is a strong positive for all countries, although the size of the coefficient is relatively small.

The SURE model shows that the impact of hosting the ATP tennis tournaments on tourism demand is significant for the United Kingdom and Canada but not for Australia. Though sport events usually last several days, we quantify the hypothetical per day contribution of the event. We find that hosting a one-day ATP tennis tournament in the United Kingdom corresponds to 0.4% increase in tourist inflow and 0.7% in Canada, compared to a month without such events, holding everything else constant. The PGA Golf, on the other hand, contributes 2% increase in tourist inflow in Australia and 4.2% in the United Kingdom. The largest increase comes from F1 in Canada with 14.3%, compared to a month without such events. Due to the limitations of our data, what we can confidently show is how the different sport events associate with the tourist arrivals in that particular month by controlling for other possible explanations that are well-known factors of tourist arrivals in the literature.

For the cross-country comparison (same sport), we use the Wald test to examine the equality of coefficients of hosting sports events across the three models. The test statistics are 26.23 (TENNIS), 0.97 (F1) and 27.4 (GOLF), respectively. The test statistics for Tennis and Golf are significant at the 1% significance level, suggesting that countries do not receive the same level of tourist inflow from hosting the same sporting event.⁷ For the within-country comparison (different sports), we find that the differences among sports are even more divergent. The Wald tests that examine the equality of coefficients across sports are 49.12 (Canada), 22.09 (United Kingdom) and 4.70 (Australia), respectively, and significant at the 1% level for the case of Canada and the United Kingdom. This suggests that the choice of sports to host can lead to different outcomes in Canada and the United Kingdom but less so for Australia. By evaluating the size and significance of the coefficients of sports variables in Table 4, we note that Australia’s tourism demand has not benefited much from hosting international sports events. Among all the statistically significant sports events, the F1 Grand Prix corresponds to the largest expected increase in tourist arrivals (Canada), while the ATP tennis tournaments have the least expected increase (UK).

Our results are consistent with the Shanghai case where Huang et al. (2014) compared F1, ATP Tour Masters 1000 and Shanghai International Marathon and found the economic impact of F1 to be nearly three times that of ATP and nine times that of the marathon. In this context, the potential economic impact of hosting various sports differs from one sport to another and, further, one country to another. Additionally, our analysis finds no significant effect from hosting the F1 for Australia and the United Kingdom, whereas tennis and golf have significant positive effects for two of the three countries.

Conclusion

In this study, we conduct a comparative analysis of three annual international sporting events across three countries to determine the potential economic impact of hosting the sport, in particular international tourist inflow. Using monthly data spanning 1995 to 2018 and OLS and SURE modelling, we compare the performance of each sport within a country as well as across countries. Our results generally show that the hosting of events, and in particular sporting events, is a useful strategy to attract international tourists. However, based on the SURE model, we find that the impact of hosting the selected sporting tournaments differs from country to country and from sports to sports. In other words, in each country, the annual sporting event that shines is different. An additional day of F1 brings in an astounding 14.3% increase in tourism demand (compared to a normal month without any events) in Canada, while tennis has a much lower estimated potential impact at less than 1%. Our result is consistent with Sylt (2016) that the F1 GP in Montreal is perhaps the largest tourist event in Canada, generating more than US$90 million in spending by visitors in the Greater Montreal area. The F1 however is not a significant attractor of international tourists in the other two countries. In Australia, PGA Golf is the only significant sporting event although Tourism Victoria (2011) showed that about 10% of the attendees of the 2011 F1 in Melbourne were international visitors and that the GP provides significant branding and positioning for the city of Melbourne. For the United Kingdom, F1 is not significant but PGA Golf brings in 4.2% more tourists, above the 0.4% increase from hosting the ATP Tennis tournaments. These results indicate that there is no ‘one-size-fits-all’ rule for sports tourism. Although holding tennis and golf tournaments may lead to an increase in tourist numbers for at least two countries, it is not the optimal event for all.

Our analysis makes two important contributions. First, we study how hosting different international sporting events can attract tourists to a country. Our results show that no two sports have the same potential economic impact. For the US$6.8 billion, a year sport tourism industry in Canada,⁸ for instance, a 14.3% increase in foreign tourists month-on-month from an additional day of the F1 Grand Prix would take the country’s sports tourism industry to an even higher level, compared to hosting a PGA Golf tournament. Second, our findings suggest that each sport event leads to different results in different countries. It is unlikely that a country can enjoy success by copy–paste sporting events that are popular in another country. According to the UN World Tourism Organization, in 2016, sports tourism accounts for as much as 55% of total tourism receipts in Australia (Macintosh et al., 2019), yet we find that hosting PGA Golf is not very beneficial for boosting tourism demand in Australia, compared to hosting the same event in the United Kingdom. These contributions lead to an important implication to policy-makers. Hosting international sporting events can act as an attractive feature of a country to appeal to international tourists. The COVID-19 pandemic has been a major setback for many internationally staged events, and it will likely take a few years to recover from this major international setback. When the pandemic subsides and international travel resumes, governments will be scrambling to revive their respective tourism industry. Hosting large-scale events, including sports, will be high on their agenda. However, when deciding which major annual sporting event to host, it is best to examine ones that would give the biggest bang-for-the-buck. The F1 races, which attract nearly the same teams to every race, may be more attractive to visitors, compared to say, ATP tennis that might not feature leading players in every tournament. Although the policy-maker may consider hosting a range of international sporting events, allocation of resources to the various host organizations should take into account the likely economic impact based on a cost–benefit analysis. Hosting a tennis tournament, for instance, could result in a lower social cost compared to an F1 city circuit, if one considers the congestion and inconvenience caused to local residents, although the ability of F1 to attract international visitors may be high (Ramasamy and Yeung, 2020).

In this study, we only considered three sports and three countries. Future studies could consider other popular sports like the international marathons, cycling races and athletics as well as more countries, particularly developing countries, to further confirm the findings of the present study. Furthermore, in our study, we did not dwell into reasons why a particular sport is able to attract more tourists. The sports differ in terms of the number of fans and accessibility. For instance, tennis courts are far more accessible than an F1 circuit allowing tennis fans to be active players compared to the passive F1 enthusiast. In addition, why is it, one may ask, that the proportionate increase in international tourists to the Grand Slam Tennis in Melbourne, Australia, is much lower than the one in Wimbledon, England? Similarly, what makes the British Open Golf the most attractive tournament across the three locations and across the three sports within the United Kingdom?

Vierhaus (2018) argued that the ‘Olympic effect’ on tourism demand could last up to 20 years as the destination image of the host city/country continues to attract foreign visitors. The impact of an annual sporting event over time is yet another area of research that deserves further study. Impulse response functions, for instance, can trace the effect of a shock to an innovation and how tourism demand respond to such a shock, both in the current and future periods. In the context of the current study, it is possible to focus on the responses of tourism demand to shocks from prolonging the sports event. A comparison of the long-term impact of these annual sporting events and contrasting it with mega sporting events like the Olympics would be worthwhile extensions to our current work.

Although these questions also form some of the limitations of the present study, we hope it will open new avenues for further study as these annual sporting events would benefit more countries than the few that get the opportunity to host mega sporting events once in a lifetime.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Howei Wu

Matthew Yeung

Notes

Author biographies

Professor Bala Ramasamy is Professor of Economics at China Europe International Business School (CEIBS). His research interest focuses on Asian economies, Foreign Direct Investment, Corporate Social Responsibility and International Business Strategy. His research has been published in the Journal of Business Ethics, World Economy, the Journal of World Business, the Journal of World Investment and Trade, the Journal of Business Research, Tourism Analysis, among others. His views are regularly sought by the media. He has been interviewed by CCTV, Bloomberg, AP, Al Jazeera, Channel NewsAsia, etc. His comments have also appeared in the Wall Street Journal, China Daily, Global Times, Diplomat and other newspaper around the world.

Dr Howei Wu is a Lecturer in Economics at China Europe International Business School (CEIBS). Prior to joining CEIBS, she was an Assistant Professor and Economist for the China Macroeconomic Analysis and Forecast Project in Institute for Advanced Research at Shanghai University of Finance and Economics. Dr Wu received her PhD in Economics from Stanford University, BA in Economics from Tsinghua University (Taiwan) and B.S in Quantitative Finance from Tsinghua University (Taiwan). Her research interest lies in Macroeconomics, specializing in monetary economics and alternatives to rational expectations. She also focuses on innovation policies, tourism economics and the development of China’s economy.

Dr Matthew Yeung is Associate Professor of Research Methods at the Institute of International Business and Governance of Hong Kong Metropolitan University. He received his PhD from the University of Nottingham. His research interest focuses on international trades, foreign direct investment, consumer sentiment and big data analysis. His research has been published in the Journal of Business Ethics, World Economy, the Journal of World Business, the Journal of World Investment and Trade, the Journal of Business Research, Current Issues in Tourism, among others.

References

Allen

Yap

(2009) Modelling Australian Domestic Tourism Demand: A Panel Data Approach. Perth, Australia: Edith Cowan University Publication.

Baade

Matheson

. (2002) “Bidding for the Olympics: Fool’s Gold?”. In: Barros

Ibrahim

and Szymanski

(eds), Transatlantic Sport. London, UK: Edward Elgar Publishing, 127–151.

Baumann

Matheson

(2018) Mega-events and tourism: the case of Brazil. Contemporary Economic Policy 36: 292–301.

Behsudi

(2020) Wish You Were Here. Finance & Development, 36–39. December 2020

Chaisumpunsakul

Pholphirul

(2018) Does international trade promote international tourism demand? Evidence from Thailand’s trading partners. Kasetsart Journal of Social Sciences 39: 292–301.

Crompton

Lee

Shuster

(2001) A guide for undertaking economic impact studies: the spring-fest example. Journal of Travel Research 40: 79–87.

Daniels

Norman

(2003) Estimating the economic impacts of seven regular sport tourism events. Journal of Sport & Tourism 8(4): 214–222.

Fiebig

(2003) Seemingly unrelated regression. In: Baltagi

(ed), A Companion to Theoretical Econometrics. Blackwell Publishing Ltd, 101–121.

Fourie

Santana-Gallego

(2011) The impact of mega-sport events on tourist arrivals. Tourism Management 32: 1364–1370.

10.

Ridderstaat

Jia

(2020) Are all tourism markets equal? Linkages between market-based tourism demand, quality of life, and economic development in Hong Kong. Tourism Management 77: 104–115.

11.

Getz

(1997), Event Management and Event Tourism. New York, NY: Cognizant Communication Corporation.

12.

Giampiccoli

Lee

Nauright

(2015) Destination South Africa: comparing global sports mega-events and recurring localised sports events in South Africa for tourism and economic development. Current Issues in Tourism 18(3): 229–248.

13.

Gibson

Kaplanidou

Kang

(2012) Small-scale event sport tourism: A case study in sustainable tourism. Sport Management Review 15(2): 160–170.

14.

Görmüş

Göçer

(2010) The socio-economic determinant of tourism demand in Turkey: a panel data approach. International Research Journal of Finance and Economics 55: 88–99.

15.

Green

Chalip

(1998) Sport tourism as the celebration of subculture. Annals of Tourism Research 25(2): 275–291.

16.

Gruben

Moss

(2012) Do the olympics create sustained increases in international tourism? Journal of International Business Research 11(1): 135–148.

17.

Gunn

(1972), Vacationscape: Designing Tourist Regions. Washington, USA: Taylor & Francis.

18.

Hall

(1992), Hallmark Tourist Events: Impacts, Management and Planning. London, UK: Belhaven Press.

19.

Hallmann

Zehrer

Müller

(2015) Perceived destination Image. Journal of Travel Research 54(1): 94–106.

20.

Hamilton

Maddison

Tol

RSJ

(2005) Climate change and international tourism: a simulation study. Global Environmental Change 15(3): 253–266.

21.

Hinch

Higham

JES

(2001) Sport tourism: a framework for research. International Journal of Tourism Research 3: 45–58.

22.

Wall

(2005) Environmental management, environmental image and the competitive tourist attraction. Journal of Sustainable Tourism 13(6): 617–635.

23.

Huang

Mao

Kim

S-K

, et al. (2014) Assessing the economic impact of three major sport events in China: the perspective of attendees. Tourism Economics 20(6): 1277–1296.

24.

Humphreys

(2006) The economic impact of sporting facilities. In: Andreff

Szymanski

(eds), Handbook on the Economics of Sport. Cheltenham, England: Edward Elgar Publishing.

25.

Joshi

(2016) Why the Herald Sun Is Australia’s Most-Read Newspaper. MediaWeek. Available at: https://mediaweek.com.au/herald-sun-most-read-newspaper/ (accessed 30 May 2019).

26.

Kaplanidou

Jordan

Funk

, et al. (2012) Recurring sport events and destination image perceptions: impact on active sport tourist behavioral intentions and place attachment. Journal of Sport Management 26: 237–248.

27.

Kaplanidou

Gibson

(2010) Predicting behavioral intentions of active event sport tourists: The case of a small-scale recurring sports event. Journal of Sport & Tourism 15(2): 163–179.

28.

Kim

S-K

Park

J-A

, et al. (2017) Measuring the economic impacts of major sports events: the case of Formula One Grand Prix (F1). Asia Pacific Journal of Tourism Research 22(1): 64–73.

29.

Kosnan

SSA

Ismail

(2012) Demand factors for international tourism in Malaysia : 1998-2009. Prosiding Perkem VII, Jilid 1(1): 44–50.

30.

Kusni

Kadir

Nayan

(2013, International tourism demand in Malaysia by tourists from OECD countries: a panel data econometric analysis. Procedia Economics and Finance 7: 28–34.

31.

Könecke

Preuss

Schütte

(2017) Direct regional economic impact of Germany’s 1. FC Kaiserslautern through participation in the 1. Bundesliga. Soccer & Society 18(7): 988–1011.

32.

Lise

Tol

RSJ

(2002) Impact of climate on tourist demand. Climatic Change 55: 429–449.

33.

Macintosh

Bravo

(2019), International Sports Management. Illinois, USA: Human Kinetics.

34.

Martins

Gan

Ferreira-Lopes

(2017) An empirical analysis of the influence of macroeconomic determinants on World tourism demand. Tourism Management 61: 248–260.

35.

Mitchell

Stewart

(2015) What should you pay to host a party? An economic analysis of hosting sports mega-events. Applied Economics 47(15): 1550–1561.

36.

NADBank (2013) Toronto star remains country’s most-read paper. Available at: https://www.thestar.com/business/2014/04/22/toronto_star_remains_countrys_mostread_paper.html (accessed 30 May 2019).

37.

Naudé

Saayman

(2005) Determinants of tourist arrivals in Africa: A panel data regression analysis. Tourism Economics 11(3): 365–391.

38.

Nielsen Sports (2018) World Football Report 2018. The Nielsen Company. https://www.nielsen.com/wp-content/uploads/sites/3/2019/04/world-football-report-2018.pdf.

39.

Nogawa

Yamaguchi

Hagi

(1996) Assessing the tourism benefits of an open access sports tourism in sports-for-all events: Case studies of a single-night event and a multiple-night event. Journal of Travel Research 35 (2): 45–55.

40.

Oldenbloom

(2006) Costs and Benefits of Major Sports Events. Amsterdam, The Netherlands: MeerWaarde.

41.

Papanikos

(2015) The economic effects of a marathon as a sport tourism event. Athens Journal of Sports 2(4): 225–239.

42.

Pham

Nghiem

Dwyer

(2017) The determinants of Chinese visitors to Australia: a dynamic demand analysis. Tourism Management 63: 268–276.

43.

Preuss

(2006) Olympics. In: Andreff

Szymanski

(eds), Handbook on the Economics of Sport. Cheltenham, England: Edward Elgar Publishing Ltd, 183–196.

44.

Preuss

(2007) The conceptualisation and measurement of mega sport event legacies. Journal of Sport and Tourism 12(3): 207–228.

45.

Ramasamy

Yeung

MCH

(2020) Does Formula-1 Grand Prix sattract tourists? Tourism Analysis 25(4): 473–479.

46.

Ritchie

(1984) Assessing the impact of hallmark events: conceptual and research issues. Journal of Travel Research 23(2): 2–11.

47.

Roche

Spake

Joseph

(2013) A model of sporting event tourism as economic development. Sport, Business and Management: An International Journal 3(2): 147–157.

48.

Rojas-Méndez

Davies

Jamsawang

, et al. (2019) Explaining the mixed outcomes from hosting major sporting events in promoting tourism. Tourism Management 74: 3000–3009.

49.

Schulz

(2010) Introduction in Player and Referee: Conflicting Interests and the 2010 FIFA World Cup. Cape Town, South Africa: Institute for Security Studies.

50.

Seetaram

(2010) Use of dynamic panel cointegration approach to model international arrivals to Australia. Journal of Travel Research 49(4): 414–422.

51.

Solberg

Preuss

(2007) Major sport events and long-term tourism impacts. Journal of Sport Management 21(2): 215–236.

52.

Song

(2008) Tourism demand modelling and forecasting-a review of recent research. Tourism Management 29(2): 203–220.

53.

Song

Witt

, et al. (2010) Tourism demand modelling and forecasting: how should demand be measured? Tourism Economics 16(1): 63–81.

54.

Sylt

(2016) Driving tourism: how Montreal uses F1 for promotion. Forbes. Available at: https://www.forbes.com/sites/csylt/2016/06/13/driving-tourism-how-montreal-uses-f1-for-promotion/#68b2a7731d8c (accessed 13 June 2016).

55.

Sylt

(2017) The $1 billion cost of hosting an F1 race. Forbes. Available at: https://www.forbes.com/sites/csylt/2017/03/13/the-1-billion-cost-of-hosting-an-f1-race/#5bdba2f64f79 (accessed 13 March 2017).

56.

Taks

Chalip

Green

(2015) Impacts and strategic outcomes from non-mega sport events for local communities. European Sport Management Quarterly 15(1): 1–6.

57.

Thomson

Cuskelly

Toohey

, et al. (2019) Sport event legacy: a systematic quantitative review of literature. Sport Management Review 22: 295–321.

58.

Vierhaus

(2018) The international tourism effect of hosting the olympic games and the FIFA world cup. Tourism Economics 25(7): 1009–1028.

59.

Victoria

(2011) Formula One Australian Grand Prix: Benefits to Victoria. Victoria. Available at: https://www.business.vic.gov.au/__data/assets/pdf_file/0003/1551909/tourism_australian_grand_prix_report_2011.pdf (accessed 10 May 2019).

60.

Walker

(2019) Mail Overtakes Sun as UK’s Most-Read Newsbrand, Pamco Readership Figures Show. Press Gazette. Available at: https://www.pressgazette.co.uk/mail-overtakes-sun-as-uks-most-read-newsbrand-pamco-readership-figures-show/ (accessed 19 June 2019).

61.

Wan

Song

(2019) Economic impact assessment of mega-events in the United Kingdom and Brazil. Journal of Hospitality & Tourism Research 43(7): 1044–1067.

62.

Wang

GHK

Hidiroglou

Fuller

(1980) Estimation of seemingly unrelated regression with lagged dependent variables and autocorrelated errors. Journal of Statistical Computation and Simulation 10(2): 133–146.

63.

Witt

(1995) Forecasting tourism demand: a review of empirical research. International Journal of Forecasting 11(3): 447–475.

64.

Yusof

Omar Fauzee

Mohd Shah

, et al. (2009) Exploring small-scale sport event tourism in Malaysia. Research Journal of International Studies 9(9): 47–58.

65.

Zellner

(1962) An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association 57(298): 348–368.

66.

Ziakas

Boukas

(2016) The emergence of “small-scale” sport events in “small Island” developing states: towards creating sustainable outcomes for Island communities. Event Management 20: 537–563.

67.

Zimbalist

(2016) The organization and economics of sports mega-events. Intereconomics 51(3): 110–111.